A learning automata-based algorithm for determination of the number of hidden units for three-layer neural networks

There is no method to determine the optimal topology for multi-layer neural networks for a given problem. Usually the designer selects a topology for the network and then trains it. Since determination of the optimal topology of neural networks belongs to class of NP-hard problems, most of the existing algorithms for determination of the topology are approximate. These algorithms could be classified into four main groups: pruning algorithms, constructive algorithms, hybrid algorithms and evolutionary algorithms. These algorithms can produce near optimal solutions. Most of these algorithms use hill-climbing method and may be stuck at local minima. In this article, we first introduce a learning automaton and study its behaviour and then present an algorithm based on the proposed learning automaton, called survival algorithm, for determination of the number of hidden units of three layers neural networks. The survival algorithm uses learning automata as a global search method to increase the probability of obtaining the optimal topology. The algorithm considers the problem of optimization of the topology of neural networks as object partitioning rather than searching or parameter optimization as in existing algorithms. In survival algorithm, the training begins with a large network, and then by adding and deleting hidden units, a near optimal topology will be obtained. The algorithm has been tested on a number of problems and shown through simulations that networks generated are near optimal.     

Evolutionary Multi-task Learning for Modular Knowledge Representation in Neural Networks

The brain can be viewed as a complex modular structure with features of information processing through knowledge storage and retrieval. Modularity ensures that the knowledge is stored in a manner where any complications in certain modules do not affect the overall functionality of the brain. Although artificial neural networks have been very promising in prediction and recognition tasks, they are limited in terms of learning algorithms that can provide modularity in knowledge representation that could be helpful in using knowledge modules when needed. Multi-task learning enables learning algorithms to feature knowledge in general representation from several related tasks. There has not been much work done that incorporates multi-task learning for modular knowledge representation in neural networks. In this paper, we present multi-task learning for modular knowledge representation in neural networks via modular network topologies. In the proposed method, each task is defined by the selected regions in a network topology (module). Modular knowledge representation would be effective even if some of the neurons and connections are disrupted or removed from selected modules in the network. We demonstrate the effectiveness of the method using single hidden layer feedforward networks to learn selected n-bit parity problems of varying levels of difficulty. Furthermore, we apply the method to benchmark pattern classification problems. The simulation and experimental results, in general, show that the proposed method retains performance quality although the knowledge is represented as modules.     

Detecting communities of triangles in complex networks using spectral optimization

The study of the sub-structure of complex networks is of major importance to relate topology and functionality. Many efforts have been devoted to the analysis of the modular structure of networks using the quality function known as modularity. However, generally speaking, the relation between topological modules and functional groups is still unknown, and depends on the semantic of the links. Sometimes, we know in advance that many connections are transitive, and as a consequence, triangles have a specific meaning. Here we propose the study of the modular structure of networks considering triangles as the building blocks of modules. The method generalizes the standard modularity and uses spectral optimization to find its maximum. We compare the partitions obtained with those resulting from the optimization of the standard modularity in several real networks. The results show that the information reported by the analysis of modules of triangles complements the information of the classical modularity analysis.     

Evolutionary Learning of Modular Neural Networks with Genetic Programming

Evolutionary design of neural networks has shown a great potential as a powerful optimization tool. However, most evolutionary neural networks have not taken advantage of the fact that they can evolve from modules. This paper presents a hybrid method of modular neural networks and genetic programming as a promising model for evolutionary learning. This paper describes the concepts and methodologies for the evolvable model of modular neural networks, which might not only develop new functionality spontaneously, but also grow and evolve its own structure autonomously. We show the potential of the method by applying an evolved modular network to a visual categorization task with handwritten digits. Sophisticated network architectures as well as functional subsystems emerge from an initial set of randomly-connected networks. Moreover, the evolved neural network has reproduced some of the characteristics of natural visual system, such as the organization of coarse and fine processing of stimuli in separate pathways.     

Using augmenting modular neural networks to evolve neuro-controllers for a team of underwater vehicles

The paper presents a new generative neuro-evolutionary method called augmenting modular neural networks (AMNN). As the name of the method implies, its purpose is to construct neural networks with a modular architecture. In addition to the modularity itself, neural networks evolving according to AMNN are also characterized by gradually expanding architecture. In the beginning of the evolutionary process, all networks consist of only output modules (or a single module). After some time, if the architecture of all networks is insufficient to effectively perform a task, all of them are augmented by one hidden module. In the following generations, further hidden modules are also added and this procedure is continued until some stopping criterion is satisfied. To test performance of AMNN, the method was used to evolve neuro-controllers for a team of underwater vehicles whose common goal was to capture other vehicle behaving by a deterministic strategy (predator–prey problem). The experiments were carried out in simulation, whereas their results were used to compare AMNN with neuro-evolutionary methods designed for building monolithic neural networks.     

Evolutionary fuzzy clustering and functional modular neural network-based human recognition

Computational intelligence shows its ability for solving many real-world problems efficiently. Synergism of fuzzy logic, evolutionary computation, and neural network can lead to development of a computational efficient and performance-rich system. In this paper, we propose a new approach for solving the human recognition problem that is the fusion of evolutionary fuzzy clustering and functional modular neural networks (FMNN). Evolutionary searching technique is applied for finding the optimal number of clusters that are generated through fuzzy clustering. The functional modular neural network has been used for recognition process that is evaluated with the help of integration based on combining the outcomes of FMNN. Performance of the proposed technique has been empirically evaluated and analyzed with the help of different parameters.

     

Fuzzy logic in the gravitational search algorithm enhanced using fuzzy logic with dynamic alpha parameter value adaptation for the optimization of modular neural networks in echocardiogram recognition

In this paper the main goal is to find the optimal architecture of modular neural networks, which means finding out the optimal number of modules, layers and nodes of the neural network. The fuzzy gravitational search algorithm with dynamic parameter adaptation is used for optimizing the modular neural network in a particular pattern recognition application. The proposed method is applied to medical images in echocardiogram recognition. One of the most common methods for detection and analysis of diseases in the human body, by physicians and specialists, is the use of medical images. Simulation results of the proposed approach in echocardiogram recognition show the advantages of using the fuzzy gravitational search in the optimization of modular neural networks. In this case the proposed approach provides a very good 99.49% echocardiogram recognition rate.     

强化学习在导弹制导中的应用

阐述了强化学习的基本原理和特点,讨论了强化学习中评价函数的神经网络近似问题,重点分析了采用多神经网络近似评价函数的学习问题,实现了状态空间或任务的自动分解,提高了评价函数的推广能力,网络的学习是离线进行,并作为反馈控制器在线应用,并以A－学习为例,将强化学习应用于导弹的制导问题,仿真结果表明了强化学习在导弹制导或控制问题中的应用前景和有效性。     

Handwritten digit recognition by adaptive-subspace self-organizingmap (ASSOM)

The adaptive-subspace self-organizing map (ASSOM) proposed by Kohonen is a recent development in self-organizing map (SOM) computation. In this paper, we propose a method to realize ASSOM using a neural learning algorithm in nonlinear autoencoder networks. Our method has the advantage of numerical stability. We have applied our ASSOM model to build a modular classification system for handwritten digit recognition. Ten ASSOM modules are used to capture different features in the ten classes of digits. When a test digit is presented to all the modules, each module provides a reconstructed pattern and the system outputs a class label by comparing the ten reconstruction errors. Our experiments show promising results. For relatively small size modules, the classification accuracy reaches 99.3% on the training set and over 97% on the testing set.     

A cooperative constructive method for neural networks for pattern recognition

In this paper, we propose a new constructive method, based on cooperative coevolution, for designing automatically the structure of a neural network for classification. Our approach is based on a modular construction of the neural network by means of a cooperative evolutionary process. This process benefits from the advantages of coevolutionary computation as well as the advantages of constructive methods. The proposed methodology can be easily extended to work with almost any kind of classifier.The evaluation of each module that constitutes the network is made using a multiobjective method. So, each new module can be evaluated in a comprehensive way, considering different aspects, such as performance, complexity, or degree of cooperation with the previous modules of the network. In this way, the method has the advantage of considering not only the performance of the networks, but also other features.The method is tested on 40 classification problems from the UCI machine learning repository with very good performance. The method is thoroughly compared with two other constructive methods, cascade correlation and GMDH networks, and other classification methods, namely, SVM, C4.5, and k nearest-neighbours, and an ensemble of neural networks constructed using four different methods.     

Multiprocessor Scheduling with the Aid of Network Flow Algorithms

In a distributed computing system a modular program must have its modules assigned among the processors so as to avoid excessive interprocessor communication while taking advantage of specific efficiencies of some processors in executing some program modules. In this paper we show that this program module assignment problem can be solved efficiently by making use of the well-known Ford–Fulkerson algorithm for finding maximum flows in commodity networks as modified by Edmonds and Karp, Dinic, and Karzanov. A solution to the two-processor problem is given, and extensions to three and n-processors are considered with partial results given without a complete efficient solution.     

基于小波网络和多模块网络的数字识别

本文研究一种新的数字识别方法,这种方法用小波神经网络抽取特征、用多模块结构神经网络作模式分类器。小波分解的函数近似能力和人工神经网络的学习能力结合起来形成的小波神经网络,有着良好的特征描述性能,可用作特征抽取工具。多模块结构的神经网络将一个k类的模式分类问题转换为k个互相独立的2类分类问题。这种结构将一个复杂的分类问题化解为多个简单的分类问题,各个模块互相并联,各自负责一种模式的识别。用这种修改过的多模块结构网络的BP训练方法,可加速训练和提高训练精度,并且各模块可互相独立地进行训练。用美国NIST数字样本进行训练及测试,结果良好。这种方法可用于更广泛的平面图形识别。     

MIGA, A Software Tool for Nonlinear System Modelling with Modular Neural Networks

This paper presents a software tool suitable for dynamic system modelling. The models generated by this tool are modular neural networks, see [1]. Each module behaves like a functional block and is connected to the other modules like in classical block diagrams. This tool allows the inclusion of a priori knowledge and, furthermore, to extract physical information from the models, once the system has learned. The modelling tool is capable of automatic model generation, parameter estimation and model validation.     

Cognitive engineering based knowledge representation in neural networks

A model of a human neural knowledge processing system is presented that suggests the following. First, an entity in the outside world lends to be locally encoded in neural networks so that the conceptual information structure is mirrored in its physical implementation. Second, the knowledge of problem solving is implemented in a quite implicit way in the internal structure of the neural network (a functional group of associated hidden neurons and their connections to entity neurons) not in individual neurons or connections. Third, the knowledge system is organized and implemented in a modular fashion in neural networks according to the local specialization of problem solving where a module of neural network implements an inter-related group of knowledge such as a schema, and different modules have similar processing mechanisms, but differ in their input and output patterns. A neural network module can be tuned just as a schema structure can be adapted for changing environments. Three experiments were conducted to try to validate the suggested cognitive engineering based knowledge structure in neural networks through computer simulation. The experiments, which were based on a task of modulo arithmetic, provided some insights into the plausibility of the suggested model of a neural knowledge processing system.     

Cognitive engineering based knowledge representation in neural networks

Abstract

A model of a human neural knowledge processing system is presented that suggests the following. First, an entity in the outside world lends to be locally encoded in neural networks so that the conceptual information structure is mirrored in its physical implementation. Second, the knowledge of problem solving is implemented in a quite implicit way in the internal structure of the neural network (a functional group of associated hidden neurons and their connections to entity neurons) not in individual neurons or connections. Third, the knowledge system is organized and implemented in a modular fashion in neural networks according to the local specialization of problem solving where a module of neural network implements an inter-related group of knowledge such as a schema, and different modules have similar processing mechanisms, but differ in their input and output patterns. A neural network module can be tuned just as a schema structure can be adapted for changing environments. Three experiments were conducted to try to validate the suggested cognitive engineering based knowledge structure in neural networks through computer simulation. The experiments, which were based on a task of modulo arithmetic, provided some insights into the plausibility of the suggested model of a neural knowledge processing system.     

LOCAL LEARNING IN NETWORKS OF UNIVERSAL ANALOGIC NEURONS

Biological neural systems exhibit the property of locality in all the calculations and structures. Classical artificial neural networks normally use an external system that performs some of the operations, mainly the learning algorithm. This arrangement means a strong dependence on external programs and machines. Learning algorithms must be implemented with local computations. Each unit has to be able to estimate its own contribution to the global error, according to the information about the errors of other units and local information. If all the modules are similar in physical connection characteristics, we can have a universal type of parametric modules. The desired final development is a general model in which all known neural network models conform. Self-programming is accomplished by means of an internal algorithm in the module. The learning is the adjustment of model parameters (indeed structural parameters). In this paper, the emphasis is on a particular case to illustrate the possibilities of inserting learning into the modules forming the network.     

有向图中模块化航天器系统相对轨道的自适应分布式一致性

基于多智能体系统一致性理论,对模块化航天器相对轨道的分布式一致性问题进行了研究.各模块之间的信息交互拓扑结构为更具一般性的有向图.当存在模块质量不确定性的情形下,设计了仅依赖模块自身及其邻近模块信息且无需模块间相对速度信息的自适应控制算法.针对模型中存在外部干扰的情形,通过引入带有时变自适应参数的变结构控制项,实现了对未知上界干扰的补偿,并且证明了闭环系统是渐近稳定的.此外,本文所设计的算法具有分布式的特点,不会因为模块数量的增多而增加所提算法的复杂度.最后对6个模块组成的模块化航天器的编队飞行进行了仿真分析,仿真结果表明本文设计的控制律是有效可行的.     

Pipelined recurrent fuzzy neural networks for nonlinear adaptive speech prediction.

A class of pipelined recurrent fuzzy neural networks (PRFNNs) is proposed in this paper for nonlinear adaptive speech prediction. The PRFNNs are modular structures comprising a number of modules that are interconnected in a chained form. Each module is implemented by a small-scale recurrent fuzzy neural network (RFNN) with internal dynamics. Due to module nesting, the PRFNNs offer a number of desirable attributes, including decomposition of the modeling task, enhanced temporal processing capabilities, and multistage dynamic fuzzy inference. Tuning of the PRFNN adaptable parameters is accomplished by a series of gradient descent methods with different weighting of the modules and the decoupled extended Kalman filter (DEKF) algorithm, based on weight grouping. Extensive experimentation is carried out to evaluate the performance of the PRFNNs on the speech prediction platform. Comparative analysis shows that the PRFNNs outperform the single-RFNN models in terms of the prediction gains that are obtained and computational efficiency. Furthermore, PRFNNs provide considerably better performance compared to pipelined recurrent neural networks, for models with similar model complexity.     

基于集成神经网络入侵检测系统的研究与实现

为解决传统入侵检测模型所存在的检测效率低,对未知的入侵行为检测困难等问题,对集成学习进行了研究与探讨,提出一种采用遗传算法的集成神经网络入侵检测模型,阐述了模型的工作原理和各模块的主要功能.模型通过遗传算法寻找那些经过训练后差异较大的神经网络进行集成.实验表明,集成神经网络与检测率最好的单个神经网络相比检测率有所提高.同时,该模型采用机器学习方法,可使系统能动态地适应环境,不仅对已知的入侵具有较好的识别能力,而且能识别未知的入侵行为,从而实现入侵检测的智能化.